Breast cancer is the second most common cancer among U.S. women with about 183,000 cases annually and an estimated 46,000 deaths in 1995. The purpose of this project is to identify novel chemopreventive agents active against breast cancer in animal models and eventually develop a clinically acceptable regimen for prevention of breast cancer in human populations at high risk. We have induced breast carcinogenesis in 55-day old virgin Sprague-Dawley rats by initiation with N-methyl-nitrosourea (NMU). We have investigated the efficacy, potency and toxicity of various potential chemopreventive agents used either alone or in combination by mixing them in the diet. We have focused on compounds classified broadly as terpenes and including both known ligands for the steroid receptor superfamily, such as retinoids, deltanoids, and classical steroids, as well as triterpenoids (such as ursolic acid) for which receptors are as yet unknown. We have shown the efficacy of 9-cis-retinoic acid (RA), N-ethyl-9-cis-retinamide, N-(4-Hydroxyphenyl)-9-cis- retinamide, N-(4-, 3-, or 2-hydroxyphenyl) retinamide and vitamin D analog Ro24-5531 alone and in combination with estrogen modifiers tamoxifen or raloxifene in preventing breast cancer. Ro24-5531, all-trans-RA and 9-cis-RA inhibited the growth of differentiated but not undifferentiated breast cancer cell lines. We have also studied effects of the triterpenoids such as ursolic acid, on both cells in vitro, and in combination with retinoids and/or estrogen response modifiers in prevention of mammary carcinogenesis. Overall, the studies suggest that effectiveness is maximized and toxicity minimized by combined use of several different chemopreventive agents which act by distinct mechanisms, and that clinical evaluation of the combination of these agents for breast cancer chemoprevention or for adjuvant therapy in women at high risk of developing breast cancer should be considered.